Keystone bacteria dynamics in chronic obstructive pulmonary disease (COPD): Towards differential diagnosis and...

Preventing exacerbations in Chronic Obstructive Pulmonary Disease (COPD) is crucial due to the high mortality rate and the associated costs of hospitalization for patients during exacerbations. Despite the proven influence of the lung microbiome on disease control, the dynamics of bacterial communication in different stages of COPD remain unknown. This study aimed to propose a group of candidate bacteria for the differential diagnosis of different states of COPD based on the relative abundance correlation of bacteria in lung sputum samples. We compared microbiome data collected from 101 COPD patients in stable and exacerbation states, as well as 124 healthy controls from two separate general cohorts, to determine the major microbiome and keystone genera. To validate our findings, we utilized two additional distinct public datasets, each comprising 81 healthy subjects and 87 COPD patients in stable condition, exacerbation, and post-treatment phases. During COPD exacerbation, Porphyromonas, Clostridium, Moryella, and Megasphaera were identified as phenotype-specific keystone genera, while Prevotella, Streptococcus, Haemophilus, and Veillonella were consistently present across all datasets as core microbiome members. Changes in keystone genera during different COPD stages indicate rewiring of bacterial interactions, with increased keystone bacteria and network connectivity observed during dysbiosis and more severe COPD. Bifidobacterium showed probiotic potential, positively correlating with Lactobacillus during exacerbation, while Neisseria and Haemophilus increased in abundance, and negatively correlated with key probiotic bacteria. These findings indicate promising potential for the simultaneous use of Bifidobacterium along with Lactobacillus as a therapeutic candidate to prevent COPD exacerbations in lung health, underscoring the need for further research in future clinical studies.